Xenobiotic shock experiments were conducted on lab-scale continuous-flow activated sludge systems to examine activated sludge treatment performance and to determine the xenobiotic degrader loss after periods of xenobiotic absence. The systems were operated with normal influent of a xenobiotic and a biogenic substrate until steady state, and were then artificially disturbed by removing and re-adding the xenobiotic in the influent. Substantial xenobiotic leaks were found when xenobiotic absent time was approximately one mean cell residence time (0(c)), and the system failed when xenobiotic absent time was longer than a 0(c). Amount of degrader at the time of dual substrate steady state was estimated to be approximately 6% of the total sludge. As the xenobiotic absence time was lengthened, degrader amount in the system was reduced exponentially at a half life of approximately three days. The loss rate could be attributed mainly to the rate of displacement by 0(c) operation, followed by endogenous decay and de-acclimation loss. (C) 2008 Elsevier Ltd. All rights reserved